ession rare action compression
Ultrasound in synthetic organic chemistry
UNSTABLE SIZE COLLAPSE
Timothy J. Mason
Sonochemistry Centre, School of Natural and Environmental Sciences, Coventry University, Coventry,
UK CV1 5FB
High-power ultrasound can generate cavitation within a by a European Society in 1990 and then other national groups
liquid and through cavitation provide a source of energy has meant that the subject has expanded greatly over the last few
which can be used to enhance a wide range of chemical years.
processes. Such uses of ultrasound have been grouped under There are a range of applications for the uses of ultrasound in
the general name sonochemistry. This review will concen- chemistry which include synthesis, environmental protection
trate on applications in organic synthesis where ultrasound (the destruction of both biological and chemical contaminants)
seems to provide a distinct alternative to other, more and process engineering (improved extraction, crystallisation,
traditional, techniques of improving reaction rates and electroplating and new methods in polymer technology).
product yields. In some cases it has also provided new
synthetic pathways. 2 Fundamental aspects
Ultrasound is defined as sound of a frequency beyond that to
which the human ear can respond. The normal range of hearing
is between 16 Hz and about 18 kHz and ultrasound is generally
The use of ultrasound in chemistry (sonochemistry) offers the considered to lie between 20 kHz to beyond 100 MHz.
synthetic chemist a method of chemical activation which has Sonochemistry generally uses frequencies between 20 and 40
broad applications and uses equipment which is relatively kHz because this is the range employed in common laboratory
inexpensive. The driving force for sonochemistry is cavitation equipment. However since acoustic cavitation can be generated
and so a general requirement is that at least one of the phases of well above these frequencies, recent researches into sonochem-
the reaction mixture should be a liquid. When laboratory istry use a much broader range (Fig. 1). High frequency
research in sonochemistry began it seemed to be mainly a ultrasound from around 5 MHz and above does not produce
method of initiating intransigent reactions especially those cavitation and this is the frequency range used in medical
which depended upon the activation of metallic or solid imaging.
reagents. Its development in the past 15 years however has
revealed that it has far wider applicability than this and also that 0 10 102 103 104 105 106 107
it presents a significant scientific challenge to understanding its
underlying physical phenomenon—acoustic cavitation. The
ever expanding number of applications of sonochemistry in
synthesis has made the subject attractive to many experimen-
talists and interest has spread beyond academic laboratories into
industry and chemical engineering.1–5 Human hearing 16Hz-18kHz
It was in 1986 that the first ever International Symposium on
Sonochemistry was held at Warwick University UK as part of Conventional power ultrasound 20kHz-40kHz
the Autumn Meeting of the Royal Society of Chemistry.6 This
meeting was significant in that it was the beginning of serious Range for sonochemistry 20kHz-2MHz
interest in the uses of ultrasound in chemistry as a study in itself.
Diagnostic ultrasound 5MHz-10MHz
Of course sonochemistry dates back much further than this. Its
origins can be traced to the early part of this century with the
Fig. 1 Sound frequency ranges
discoveries of echo sounding and the mechanical use of power
ultrasound for emulsification. The formation of the Royal
Society of Chemistry Sonochemistry Group in 1987 followed Like all sound energy, ultrasound is propagated via a series of
compression and rarefaction waves induced in the molecules of
the medium through which it passes. At sufficiently high power
the rarefaction cycle may exceed the attractive forces of the
Professor Mason obtained a molecules of the liquid and cavitation bubbles will form. These
BSc (1967) in chemistry and bubbles will grow over a few cycles taking in some vapour or
PhD (1970) in organic chem- gas from the medium (rectified diffusion) to an equilibrium size
istry from Southampton Uni- which matches the frequency of bubble resonance to that of the
versity. After periods at Am- sound frequency applied. The acoustic field experienced by the
herst College, USA, York Uni- bubble is not stable because of the interference of other bubbles
versity and Bradford University forming and resonating around it. As a result some bubbles
he joined Coventry Polytechnic suffer sudden expansion to an unstable size and collapse
(now University) in 1975. He is violently. It is the fate of these cavities when they collapse
currently chairman of the RSC which generates the energy for chemical and mechanical effects
Sonochemistry group and Pres- (Fig. 2). There are several theories which have been advanced to
ident of the European Society of explain the energy release involved with cavitation of which the
Sonochemistry and was most understandable in a qualitative sense is the ‘hot spot’
awarded a DSc in 1996. approach. Each cavitation bubble acts as a localised micro-
reactor which, in aqueous systems, generates temperatures of
Chemical Society Reviews, 1997, volume 26 443
3.1 The ultrasonic cleaning bath
The simple ultrasonic cleaning bath is by far the most widely
available and cheapest source of ultrasonic irradiation for the
rarefaction compression rarefaction compression
chemical laboratory. Although it is possible to use the bath itself
as a reaction vessel this is seldom done because of problems
associated with corrosion of the bath walls and containment of
any evolved vapours and gases. The normal usage therefore
involves the immersion of standard glass reaction vessels into
the bath which provides a fairly even distribution of energy into
ONE CYCLE the reaction medium (Fig. 4). The reaction vessel does not need
5000 °C any special adaptation, it can be placed into the bath, thus an
2000 atm inert atmosphere or pressure can be readily maintained
throughout a sonochemical reaction. The amount of energy
BUBBLE BUBBLE GROWS IN REACHES VIOLENT
FORMS SUCCESSIVE CYCLES UNSTABLE SIZE COLLAPSE which reaches the reaction through the vessel walls is low—
BY RECTIFIED DIFFUSION normally between 1 and 5 W cm22. Temperature control in
commercial cleaning baths is generally poor and so the system
Fig. 2 Sound propagation in a liquid showing cavitation bubble formation may require additional thermostatic control.
several thousand degrees and pressures in excess of one
thousand atmospheres. reaction mixture
water + detergent
In addition to the generation of extreme conditions within the
bubble there are also major mechanical effects produced as a
result of its rapid collapse. These are also of significance in
synthesis and include very rapid degassing of the cavitating stainless
liquid (in the rarefaction cycle the newly formed bubbles will steel tank
fill with gas and be expelled from the liquid) and rapid optional
crystallisation (brought about through seed crystal generation heater
3 Laboratory equipment
The first requirement for sonochemistry is a source of transducers
ultrasound and whatever type of commercial instrument is used bonded to base
the energy will be generated via an ultrasonic transducer—a Fig. 4 The ultrasonic cleaning bath in sonochemistry
device by which mechanical or electrical energy can be
converted to sound energy. There are three main types of
ultrasonic transducer used in sonochemistry: liquid-driven 3.2 The ultrasonic probe
(effectively liquid whistles), magnetostrictive (based on the This apparatus allows acoustic energy to be introduced directly
reduction in size of certain metals, e.g. nickel, when placed in a into the system rather than rely on its transfer through the water
magnetic field) and piezoelectric. Most of the current equip- of a tank and the reaction vessel walls (Fig. 5). The power of
ment used for sonochemistry utilises transducers constructed of such systems is controllable and the maximum can be several
piezoelectric ceramics. These are brittle and so it is normal hundred W cm22. The probe system is more expensive than the
practise to clamp them between metal blocks for protection. The bath and it is slightly less convenient in use because special
overall structure is known as a piezoelectric ‘sandwich’. seals will be needed if the horn is to be used in reactions which
Usually two ceramic elements are combined so that their overall involve reflux, inert atmospheres or pressures above (or below)
mechanical motion is additive (Fig. 3). Piezoelectric trans- ambient.
ducers are very efficient and, depending on their dimensions,
can be made to operate over the whole ultrasonic range.
back mass upper (fixed)
securing bolt of metal horn
at null point
electrical contact detachable horn
front mass replacable
of metal tip
Fig. 5 The ultrasonic probe system in sonochemistry
Fig. 3 Construction of a piezoelectric sandwich transducer
4 An attempt to formulate some rules governing
The two most common sources of ultrasound for laboratory
sonochemistry are the ultrasonic cleaning bath and the ultra-
sonic horn or probe system.7 These generally operate at One of the earliest tenets of sonochemistry was that it is
frequencies of around 40 and 20 kHz, respectively. particularly good at assisting reactions involving solid reagents.
444 Chemical Society Reviews, 1997, volume 26
This is generally but not exclusively correct. A number of enter the bubble and so should be volatile. The ‘concentration’
groups are attempting to gain an understanding of the of cavitation bubbles produced by sonication using conven-
underlying principles of sonochemistry in order to be able to tional laboratory equipment is very small and so overall yields
predict which type of reaction would be most susceptible to in this type of reaction are low. Thus in the sonication of water
sonication. As a result of these efforts some guidelines have small quantities of OH· and H· radicals are generated in the
been identified. An empirical classification of sonochemical bubble and these undergo a range of subsequent reactions
reactions into three types was proposed by J.-L. Luche and was including the generation of H2O2. The highly oxidising HO·
based upon the purely chemical effects induced by cavitation.8 species can react with other moieties in the bubble or migrate to
Other (mechanical) effects of cavitation bubble collapse (e.g. the bulk solution where they have only transient existence. Such
emulsification) were considered to be physical rather than radicals can have a significant effect on both biological and
chemical and judged to be ‘false’ sonochemistry. These so- chemical species in aqueous solution and can be detected
called ‘false’ effects are often important and have been included chemically.9 Organic solvents will also slowly decompose on
in the following interpretation of the three original types of sonication but solvent decomposition is normally only a minor
reaction susceptible to sonochemical enhancement. contribution to any sonochemical reaction taking place in the
Type 1 Homogeneous systems which proceed via radical or
A synthetically useful reaction which takes place in the
radical-ion intermediates. This implies that sonication
collapsing bubble is the production of amorphous iron from the
is able to effect reactions proceeding through radicals
sonolysis of Fe(CO)5 (0.4 m) in decane under argon.10 Volatile
and further that it is unlikely to effect ionic reactions.
iron pentacarbonyl enters the bubble and is decomposed during
Type 2 Heterogeneous systems proceeding via ionic interme-
collapse. The fact that an amorphous (rather than crystalline)
diates. Here the reaction is influenced primarily
material is produced confirms that very high temperatures are
through the mechanical effects of cavitation such as
generated in the bubble and that extreme cooling rates are
surface cleaning, particle size reduction and improved
involved. Conventional production of amorphous iron requires
mass transfer. This is what has sometimes been
rapid cooling from the vapour to solid state of the order of 106
referred to as ‘false sonochemistry’.
K s21. Sonolytic decomposition of iron pentacarbonyl in
Type 3 Heterogeneous reactions which include a radical
pentane (a more volatile solvent) yields Fe3(CO)12 rather than
pathway or a mixed mechanism i.e. radical and ionic.
the metal indicating that the cavitation collapse is not so
Radical reactions will be chemically enhanced by
extreme in this solvent. Since this original report the study of
sonication but the general mechanical effect referred to
cavitation induced decomposition of iron and other metal
above may well still apply. If the radical and ionic
carbonyls has continued and expanded. In the case of molybde-
mechanisms lead to different products ultrasound
num hexacarbonyl the product is nanostructured molybdenum
should favour the radical pathway and this could lead
carbide which has proved to be an excellent dehydrogenation
to a switch in the nature of the reaction products.
In this article the term sonochemistry will be used to encompass Sucrose has a negligible vapour pressure and so cannot enter
any beneficial effect on synthesis induced by cavitation whether the bubble during sonication. A study of the effect of sonication
it is chemical or physical. on the rate of acid catalysed inversion of this material revealed
no appreciable effect. It is tempting to conclude from this that
4.1 Reactions which exemplify the ‘rules’ of sonochemistry has no effect on involatile materials in solution.
sonochemistry This is not entirely correct because bubble collapse produces
4.1.1 Homogeneous liquid-phase reactions very large shear forces in the surrounding liquid capable of
Any system involving a homogeneous liquid in which bubbles breaking the chemical bonding in polymeric materials dissolved
are produced is not strictly homogeneous, however, in sono- in the fluid.1 Over the last few years, increasing interest has
chemistry it is normal to consider the state of the system to been shown in this procedure since the net result of polymer-
which the ultrasound is applied. Sonochemical syntheses in chain rupture is a pair of macroradicals, which may recombine
homogeneous conditions are not extensively reported which randomly (resulting in a reduction in molar mass and possibly
suggests that cavitation is less effective in promoting reactions leading to a monodispersed system) or act as a radical site on
under these conditions. The few studies which have appeared which to polymerise another monomer added to the solution
indicate that sonochemical effects generally occur either inside (resulting in block copolymerization).
the collapsing bubble where extreme conditions are produced, Small accelerations, in the range 4–15%, have been found for
at the interface between the cavity and the bulk liquid where the the rate of acid catalysed hydrolysis of a number of esters of
conditions are far less extreme or in the bulk liquid immediately carboxylic acids.1 In the case of methyl ethanoate the effects (at
surrounding the bubble where the predominant effects will be 23 kHz) were attributed to the increased molecular motion
mechanical (Fig. 6). induced by the pressure gradients associated with bubble
IN THE CAVITY collapse. Similarly, the hydrolysis of the 4-nitrophenyl esters of
extreme conditions a number of aliphatic carboxylic acids at 35 °C showed
ultrasonically (20 kHz) induced rate enhancements which were
all in the range of 14–15% (Scheme 1). The activation energy
for the hydrolysis of each of the substrates varied considerably
with the alkyl substituent (R = Me, Et, Pri, But) on the
carboxylic acid and so the uniform increase in rate could not be
AT THE INTERFACE associated with any cavitational heating effect. Here again, the
modest sonochemical effect was considered to be the result of
R C H2O O
IN THE BULK MEDIA + HO NO2
O NO2 R C
intense shear forces
Fig. 6 Cavitation in a homogeneous liquid Scheme 1
In order for a chemical to experience the extreme conditions The effect of ultrasonic irradiation on the hydrolysis of
generated inside the cavitation bubble during collapse it must 2-chloro-2-methylpropane in mixed aqueous ethanolic solvents
Chemical Society Reviews, 1997, volume 26 445
of different compositions revealed more evidence for the showed no sign of re-agglomeration even after being allowed to
influence of mechanical effects.1 The rate enhancement induced stand for a period of 24 h.
by ultrasound (at 20 kHz) was found to increase with increase in
the alcohol content and to decrease as the reaction temperature LARGE PARTICLES SMALL PARTICLES
was raised. A maximum rate increase of 20-fold was observed fragmentation
at 10 °C in 50% (m/m) solvent composition. This composition surface
is closely coincident with the structural maximum for the binary erosion
ethanol–water solvent system. It is logical to suppose that if the
sonochemical enhancement is associated with solvent disrup-
tion then the maximum effect would be observed at this
4.1.2 Heterogeneous systems
In any heterogeneous system cavitation which occurs in the
liquid phase will be subject to the same conditions as have been
described above for homogeneous reactions. There will be a surface cavitation
due to defects violent collision
difference however when bubbles collapse at or near any
interface and this will depend upon the phases involved. Fig. 8 Cavitation in a particulate medium
If cavitation bubbles are formed at or near to any large solid
surface the bubble collapse will no longer be symmetrical. The The O-alkylation of 5-hydroxychromones is a difficult
large solid surface hinders liquid movement from that side and process probably as a result of hindrance to ionisation caused by
so the major liquid flow into the collapsing bubble will be from hydrogen bonding between the carbonyl and OH group coupled
the other side of the bubble. As a result of this a liquid jet will with some dispersion of the resulting phenoxide O2 charge.
be formed which is targeted at the surface with speeds in excess Thus, using 5-hydroxy-4-oxo-4H-1-benzopyran-2-carboxylic
of 100 m s21 (Fig. 7). The mechanical effect of this is equivalent acid ethyl ester as substrate in N-methylpyrrolidinone (NMP) a
to high pressure jetting and is the reason why ultrasound is so low yield (28%) of the O-propyl product is obtained after 1.5 h
effective in cleaning. Depending upon the conditions used this at 65 °C using 1-iodopropane and potassium carbonate as base
powerful jet can activate surface catalysis, force the impregna- (Scheme 2).7 The yield was greatly increased under sonication
tion of catalytic material into porous supports and generally
increase mass and heat transfer to the surface by disruption of OH O OR O
interfacial boundary layers. PrnI/K2 CO3
ASYMMETRIC COLLAPSE O COOEt O COOEt
improved mass transport Scheme 2
erosion (probe 20 kHz) and the scope of the reaction was expanded by
using a range of different haloalkanes (Table 1). Power
Inrush of liquid ultrasound would be expected to be effective in enhancing this
from one side of reaction via the reduction of the particle size of K2CO3 powder.
collapsing bubble This factor was investigated by first sonicating NMP containing
K2CO3 at 65 °C. The appropriate proportions of 1-iodopropane
and chromone were then added to the resulting very fine
suspension and the reaction was run under conventional
conditions. This resulted in around 90% product formation in 90
min at 65 °C, a reactivity similar to that obtained under
MICROJET FORMATION continuous ultrasonic irradiation except that the reaction was
approaching a definite limit at 90% yield. The fall-off suggests
Fig. 7 Cavitation near to a solid surface that the surface of the remaining K2CO3 had become deacti-
vated, and this was confirmed when sonication of the residual
For this reason the use of ultrasound in conjunction with mixture rapidly completed the reaction.
almost any electrochemical process will be beneficial and has
been the subject of extensive study. The subject has become Table 1 O-Alkylation of a hydroxychromone
known as sonoelectrochemistry.12 The particular advantages
which accrue include (a) degassing at the electrode surface, (b) Alkyl group Yield (stirred) (%) Yield (sonicated) (%)
disruption of the diffusion layer which reduces depletion of
electroactive species, (c) improved mass transport of ions across PrnI 28 100
the double layer and (d) continuous cleaning and activation of BunI 57 97
the electrode surfaces. All of these effects combine to provide BnBr 59 97
enhanced yield and improved electrical efficiency. a GLC yields, 90 min in NMP at 65 °C, sonication with 20 kHz probe
When the solid is particulate in nature, cavitation can produce system.
a variety of effects depending on the size and type of the
material (Fig. 8). These include mechanical deaggregation and A study of the Ullmann coupling reaction has provided
dispersion of loosely held clusters, the removal of surface evidence that the mechanical effects of surface cleaning coupled
coatings by abrasion and improved mass transfer to the surface. with an increase in surface area cannot fully explain the extent
Mechanical deagglomeration is a useful processing aid and is of the sonochemically enhanced reactivity. The reaction of
illustrated in the effect of sonication (in a bath) of titanium 2-iodonitrobenzene to give a dinitrobiphenyl using conven-
dioxide pigment in water. A powder sample made up in water tional methodology requires heating for 48 h and the use of a
consisting initially of agglomerates (volume mean diameter ca. tenfold excess of copper powder (Scheme 3). The use of power
19 mm) was rapidly broken apart ( < 30 s) to provide a limiting ultrasound affords a similar (80%) yield in a much shorter time
size of 1.6 mm particles. Furthermore, the sonicated sample (1.5 h) using only a fourfold excess of copper.7 During these
446 Chemical Society Reviews, 1997, volume 26
NO2 NO2 support, possibly by masking them through cavitationally
Cu induced cyanide absorption.
DMF/60 °C CH3
Scheme 3 stir 24 h
studies it was observed that the average particle size of the
copper fell from 87 to 25 mm but this increase in surface area
was shown to be insufficient to explain the large (50-fold) CH2Br
enhancement in reactivity produced by ultrasonic irradiation. KCN/Al2O3
The studies suggested that sonication assisted in either the
breaking down of intermediates and/or the desorption of CH3
products from the surface. An additional practical advantage CH2CN
was that sonication prevented the adsorption of copper on the sonicate 24 h
walls of reaction vessels, a common problem when using 77%
The collapse of cavitation bubbles at or near the interface of
immiscible liquids will cause disruption and mixing, resulting Scheme 5
in the formation of very fine emulsions (Fig. 9). This is
essentially a mechanical effect but it has been utilised in the The same group have reported an example of sonochemical
hydrolysis of benzoate esters where the emulsion was produced switching in a homogeneous reaction. The decomposition of
by a probe system.13 Using 10% NaOH the conventional lead tetraacetate in acetic acid the presence of styrene at 50 °C
hydrolysis (Scheme 4) under reflux, gave a very low yield after generates a small quantity of diacetate via an ionic mechanism.
90 min; however sonication at room temperature afforded near Under otherwise identical conditions sonication of the mixture
complete hydrolysis in 1 h. gives 1-phenylpropyl acetate predominantly through an inter-
mediate methyl radical which adds to the double bond (Scheme
6).15 These results are in accord with the proposition that radical
processes are favoured by sonication.
disruption OAc OAc AcO
radical mixed ionic
pathway pathway pathway
Fig. 9 Cavitation in a two phase liquid medium
Another example of sonochemical switching and is found in
the Kornblum–Russell reaction (Scheme 7). 4-Nitrobenzyl
Me Me bromide reacts with 2-lithio-2-nitro-propane via a predomi-
10% NaOH nantly polar mechanism to give, as a final product, 4-nitrobenz-
Me COOMe Me COO– aldehyde.16 An alternative SET pathway exists in this reaction
leading to the formation of a dinitro compound. Sonication
Me Me changes the normal course of the reaction and gives preferen-
Scheme 4 tially the latter compound, in amounts depending on the
irradiation conditions and the acoustic intensity.
Other applications of sonochemically induced emulsification
are in phase transfer catalysis, emulsion polymerisation and two NO2 NO2 NO2
phase enzymatic syntheses. –
stir O sonicate
+ N Li+
4.1.3 Reactions ‘switched’ by ultrasound S N2 SRN1
An extremely good way of demonstrating that sonochemistry is
different from other methods of enhancing chemical reactions is H O Br
to find specific reactions for which ultrasound has changed the NO2
product distribution. The first report of a reaction exhibiting
‘sonochemical switching’ came from Ando et al.14 The system Scheme 7
consisted of a suspension of benzyl bromide and alumina-
supported potassium cyanide in toluene as solvent (Scheme 5). A sonochemical switch has also been observed in the
The aim was to produce benzyl cyanide by nucleophilic formation of the indanone nucleus from o-allyl benzamides
displacement of the bromine by supported cyanide. Under (Scheme 8).17 The ketyl radical anion cyclizes to 2-methylin-
stirring alone the reaction provided diphenylmethane products danone and liberates an amide ion which deprotonates the allyl
via a Friedel–Crafts reaction between the bromo compound and moiety. The resulting carbanion then undergoes cyclization to
the solvent, catalysed by Lewis acid sites on the surface of the a-naphthol. Under sonication the first step of the process is
solid phase reagent. In contrast, sonication of the same accelerated and the ketyl is generated much more rapidly so that
constituents produced only the substitution product, benzyl only the cyclization to 2-methylindanone occurs.
cyanide. The explanation for this was based upon cavitation The Kolbe electrolysis of cyclohexanecarboxylate in aqueous
producing a structural change to the catalytic sites of the solid methanol generates a mixture of products in which bicyclohexyl
Chemical Society Reviews, 1997, volume 26 447
OH Br MgBr
CH3 CH2 CH CH3 CH3 CH2 CH CH3
+ Li/THF ultrasonic irradiation is able to initiate Grignard formation in
under 4 min compared with several hours using the traditional
ultrasound method involving periodic crushing of the metal.
The formation of cyclopropanes through the Simmons–Smith
reaction involving zinc dust and CH2I2 and an alkene suffers
Scheme 8 from several experimental drawbacks some of the major ones
being the sudden exotherm which occurs after an unpredictable
predominates (49%). In the presence of ultrasound (38 kHz) induction period, foaming and the difficulties in removing
however the product distribution was changed quite sig- finely divided metal from the reaction products. The conven-
nificantly reducing the yield of bicyclohexyl to only 7.7% tional method for enhancing this reaction relies upon activation
(Scheme 9).12 The major products were the result of two of the zinc metal by using it in the form of a zinc–silver or zinc–
electron processes through a cyclohexane carbocation which copper couple and/or using iodine or lithium in conjunction
gave cyclohexene (34%) by elimination and cyclohexyl methyl with the metal. The experimental difficulties have been
ether (32%) by solvent attack. A characteristic of many eliminated using a sonochemical procedure where no special
sonoelectrochemical processes is that the average cell potential activation of the zinc was required and good and reproducible
under sonication is less than that required conventionally. In this yields were obtained using zinc metal in the form of mossy rods
case a current density of 200 mA cm22 could be maintained at or foil (Scheme 11).19
a potential of 7.3 V with ultrasound compared with 8.3 V under
. The dehydrogenation of tetrahydronaphthalene to naph-
thalene using 3% Pd/C in digol under the influence of sonication
is accelerated by ultrasonic irradiation (Scheme 12).20 The
conventional thermal reaction in digol at 200 °C reached 55%
–e– conversion in 6 h (but thereafter reaction ceased) and only 17%
reaction was obtained in the same time at the lower temperature
of 180 °C. Under sonication at 180 °C the reaction reached
completion in 6 h. Pulsed ultrasound (at 50% cycle) was as
+ OCH3 +
effective as continuous sonication and even a 10% cycle gave
over 80% yield. These results offer considerable energy
two-electron products savings, particularly on processes carried out on a large scale.
Scheme 9 Pd/C
5 Some applications of ultrasound in synthesis
5.1 The activation of metals
Ultrasound can be used to accelerate reactions involving metals Surface activation is of great use in catalysis where metal
through surface activation which can be achieved in three ways powders such as nickel, which are generally poor catalysts, can
(a) by sonication during the reaction, (b) as a pre-treatment be activated by sonication before use. Normally, simple nickel
before the metal is used in a conventional reaction or (c) to powder is a reluctant catalyst for the hydrogenation of alkenes
generate metals in a different and more reactive form. yet ultrasonic irradiation offered a reactivity comparable with
A classic use of ultrasound is in the initiation and enhance- Raney nickel.21 In this case, ultrasound gave an unexpected
ment of synthetic reactions involving metals as a reagent or decrease in surface area due to aggregation of particles, with
catalyst. One such example is the preparation of a Grignard electron micrographs indicating a smoothing of the nickel
reagent. A long-standing problem associated with Grignard surface. Auger electron spectroscopy revealed an increase in the
reagent synthesis is that in order to facilitate reaction between nickel/oxygen ratio at the surface. The explanation suggested
the organic halide and the metal in an ether solvent all of the was that abrasion from interparticle collisions removes the
reagents must be dry and the surface of the magnesium must be oxide layer of the nickel giving the observed increased
clean and oxide free. Such conditions are difficult to achieve reactivity. A simple pre-sonication of 3 mm nickel in ethanol
and so many methods of initiating the reaction have been prior to use is quite capable of converting this powder from an
developed most of which rely on adding activating chemicals to extremely poor into an acceptable catalyst for the conventional
the reaction mixture. A very simple method of initiating the hydrogenation of oct-1-ene.
reaction is by sonication of the reaction mixture in an ultrasonic The reduction of metal salts to a finely divided very reactive
bath which avoids the need for the addition of chemical free metal generally involves refluxing the metal salt in THF
activators. The quantitative effects of ultrasound on the with a very active metal like potassium. The conditions for the
induction times for the formation of a Grignard reagent using production of these so-called Rieke powders can be ameliorated
magnesium turnings in various grades of ether have been using ultrasound such that equally reactive metal powders can
examined (Scheme 10).18 Using damp, technical grade ether be produced using lithium in THF at room temperature. An
448 Chemical Society Reviews, 1997, volume 26
example of the use of sonochemically generated Rieke powders system. Sonication provides such a method which has been used
is in the preparation of organosilicon compounds (Scheme in the synthesis of peptides (Scheme 16).25 The methodology is
13).22 effective using different solvent combinations (Table 3).
Cl3SiH SiCl3 BOC Gly + Phe N2H2Ph
'Rieke' Ni powder
CH2 CHCN CH3 CH (93%)
'Rieke' Ni powder
SiCl3 BOC Gly Phe N2H2Ph
Scheme 13 Scheme 16
A novel method of generated finely divided zinc metal is by Table 3 Dipeptide synthesis in an aqueous emulsiona
the use of pulsed sonoelectrochemistry using an ultrasonic horn
as the cathode.23 Normal electrolysis of ZnCl2 in aqueous Organic phase Stir Sonicate
NH4Cl affords a zinc deposit on the cathode. When the
electrolysis is pulsed at 300 ms on/off and the cathode is pulsed Diethyl ether 71 89
Light petroleum 12 62
ultrasonically at a 100 : 200 ms on/off ratio the zinc is produced
as a fine powder. This powder is considerably more active than aWater (75%) with organic solvent (25%) at 37 °C 12 h, 38 kHz ultrasonic
commercial zinc powder, e.g. in the addition of allyl bromide to bath.
benzaldehyde (Scheme 14).
Another and probably the most spectacular example of the
O H OH correct choice of optimized sonicating conditions has been
C reported for the microbial conversion of cholesterol to chol-
+ (82%) estenone (Scheme 17).26 Optimum conditions involved irradia-
Zn powder tion pulses of 2.8 W power applied for 5 s each 10 min and this
gave a 40% yield increase.
5.2 Enzymatic syntheses
An area of sonochemistry which is deserving of far greater
attention is the use of ultrasound to modify enzyme or whole Microbial
cell reactivity. High power ultrasound will break biological cell action
walls releasing the contents but it can also denature enzymes. It HO O
is therefore very important that when ultrasound is used in
conjunction with biological material the conditions of sonica-
tion must be carefully regulated.
Controlled sonication has been used to ‘stimulate’ a suspen- 5.3 Phase transfer and related reactions
sion of baker’s yeast to provide an inexpensive source of sterol The effect of cavitation on a suspended solid has been described
cyclase (Scheme 15, Table 2).24 This technique provides an above (section 4.1.2). Such effects become very important in the
case of reactions involving solid–liquid phase transfer catalysis.
The N-alkylation of indole with RBr [R = CH3(CH2)11] in
toluene at 25 °C in the presence of solid KOH produces a 19%
R yield in 3 h using tert-butylammonium nitrate (Scheme 18).
This yield is substantially improved by sonication to around
Brewer's 90% after only 80 min.27
Scheme 15 H R
enantioselective enzymatic synthesis of a sterol in gram
quantities. Significantly, sonication has no effect on the activity
of the isolated cell-free cyclase system, a result which In some cases sonochemistry can completely remove the
demonstrates how cell membrane disruption can occur without need for PTC as is the case in the generation of dichlorocarbene
damage to the contents. by the direct reaction between powdered sodium hydroxide and
chloroform at 40 °C using an ultrasonic bath.28 Under these
Table 2 Conversion of squalene oxide to sterol with baker’s yeast conditions styrene can be cyclopropanated in 96% yield in 1 h
when a combination of both sonication and mechanical stirring
Enzyme source Conversion (%) Enantiomer conversion (%) is used. Significantly the yield is much reduced to 38% in 20 h
with sonication alone because the power of the bath is not
Whole yeast 9.5 19 sufficient to disperse the solid reagent into the dense chloroform
Pre-treated yeast a 41.9 83.9 (Scheme 19).
a Presonication at 0 °C using a probe system (20 kHz) for 2 h. b Enzymatic One route to amino acids is via the synthesis of aminonitriles.
reaction at 37 °C, 12 h. The direct reaction between an aldehyde, KCN and NH4Cl in
acetonitrile leads to a mixture of products but in the presence of
When an enzyme is used in a two phase synthesis one of the alumina and sonication the reaction can be made more specific
important requirements is an efficient emulsification/mixing (Scheme 20).29 In the case of benzaldehyde the yield of the
Chemical Society Reviews, 1997, volume 26 449
PPh3+ Br– O
Room Temp. +
Cl PPh3+ Br– O
CN C CN
C H C OH H C OH H C NH2
target aminonitrile is poor under normal stirred conditions with Scheme 22
benzoin and hydroxynitrile predominating (Table 4). The
presence of alumina suspended in acetonitrile increases the Trialkylboranes are generally obtained through the stepwise
proportion of aminonitrile but the overall results make it clear reaction of borane with an alkene. With hindered alkenes
that the optimum reaction conditions require the presence of however the reaction is very slow. Sonication promotes rapid
suspended alumina together with sonication and then the yield reaction even with highly hindered substrates (Scheme 23).32
of target aminonitrile reaches 90%. Synthetic applications of this technique include the hydrobor-
ation/oxidation of vinyl groups.
Table 4 Strecker synthesis of an aminonitrilea
Conditions Cyanohydrin Benzoin Aminonitrile
Stir 38 21 6 + BH
Stir + Al2O3 19 9 64
Sonicate 45 22 23
Sonicate + Al2O3 3 7 90 Neat 99% yield 1 h ultrasonic bath (5 h normal)
a 25 °C, 38 kHz ultrasonic bath.
5.4 Miscellaneous syntheses BnO BnO
Synthetic applications of the sonolysis of iron carbonyl which HO O
lead to useful ferrilactones synthons have been described O
(Scheme 21). These are prepared easily and in good yields from
vinyl epoxides and either iron pentacarbonyl or, for conven-
ience and safety, diiron nonacarbonyl. The use of ferrilactones OTBDMS OTBDMS
together with ultrasonically assisted reactions of samarium
diodide and sodium phenylcyanide in natural product syntheses (i) 9-BBN , THF , ultrasound 89 %
have been reviewed.30 (ii) NaOH , H2O2
O Scheme 23
Sonochemistry has been used to improve a Friedel–Crafts
alkylation reaction used for the synthesis of the anti-inflamma-
tory agent ibuprofen (Scheme 24).33 When performed under
(CO)3Fe classical conditions (2 h at 25 °C) the reaction afforded only
R O 17% yield and for this reason the normal synthesis is via a less
(i) R′NH2 /Lewis acid direct route. Under the influence of ultrasound, using a cleaning
bath, but under otherwise identical conditions the yield was
CeIV (ii) CeIV
improved to 50%.
NR′ OSO2CH3 AlCl3
O O O CH3 CH COOH +
A rather difficult double Wittig reaction (Scheme 22) has
been effected with enhanced efficiency under sonication.31 The
process constitutes a novel type of annelation of an aromatic
ring when applied to o-quinones. It is possible to considerably Sonochemistry is an expanding field of study which continues
simplify experimental procedures with ultrasound which allows to thrive on outstanding laboratory results.34 Applications can
the use of bases which are insensitive to moisture or air. be found over a range of chemical systems, however it is in
450 Chemical Society Reviews, 1997, volume 26
heterogeneous reactions that sonochemical syntheses are most 9 P. Riesz, Free radical generation by ultrasound in aqueous solutions of
widely developed. The potential improvements afforded by volatile and non-volatile solutes, Advances in Sonochemistry, ed. T.
sonication suggest that all chemical laboratories nowadays J. Mason, JAI Press, London, 1991, vol. 2, p. 23.
10 K. S. Suslick, S.-B. Choe, A. A. Chichowlas and M. W. Grimstaff,
should be equipped with at least one small cleaning bath for
Nature, 1991, 353, 414.
simple trials. 11 T. H. Hyeon, M. M. Fang and K. S. Suslick, J. Am. Chem. Soc., 1996,
While an empirical understanding of the subject has taken 118, 5492.
sonochemists a long way towards predicting possible applica- 12 D. J. Walton and S. S. Phull, Sonoelectrochemistry, Advances in
tions considerable attention is currently being paid to gaining an Sonochemistry, ed. T. J. Mason, JAI Press, London, 1996, vol. 4,
understanding of what actually goes on in the collapsing bubble p. 205.
and in its immediate environment. In this area the chemists are 13 S. Moon, L. Duchin and J. V. Cooney, Tetrahedron Lett., 1979, 20,
finding very fruitful cooperation with engineers, physicists and 3917.
mathematicians making sonochemistry a truly interdisciplinary 14 T. Ando and T. Kimura, Ultrasonic organic synthesis involving non-
metal solids, Advances in Sonochemistry, ed. T. J. Mason, JAI Press,
London, 1991, vol. 2, p. 211.
Recent laboratory studies have revealed that for a few 15 T. Ando, P. Bauchat, A. Foucaud, M. Fujita, T. Kimura and H. Sohmiya,
heterogeneous reactions high speed stirring has a similar effect Tetrahedron Lett., 1991, 32, 6379.
to sonication.35 Thus in the cyclopropanation of styrene 16 M. J. Dickens and J.-L. Luche, Tetrahedron Lett., 1991, 32, 4709.
(Scheme 19) the yield can be improved from 3% with magnetic 17 J. Einhorn, C. Einhorn and J.-L. Luche, Tetrahedron Lett, 1988, 29,
stirring through 20% at 8000 rpm to 70% at 24 000 rpm. Such 2183.
results are intriguing in that they confirm the importance of 18 J. D. Sprich and G. S. Lewandos, Inorg. Chim. Acta, 1982, 76, 1241.
mass transfer in sonochemistry and could suggest that high 19 O. Repic and S. Vogt, Tetrahedron Lett., 1982, 23, 2729.
speed stirring involves hydrodynamic cavitation. Unlike sonica- 20 T. J. Mason, J. P. Lorimer, L. Paniwnyk, P. W. Wright and A. R. Harris,
J. Catal., 1994, 147, 1.
tion however stirring at such very high speeds is unlikely to
21 K. S. Suslick and D. J. Casadonte, J. Am. Chem. Soc., 1987, 109,
become a viable prospect in industry. 3459.
Whatever the real laws of sonochemistry might be it is clear 22 W. L. Parker, P. Boudjouk and A. B. Rajkumar, J. Am. Chem. Soc.,
that sonochemistry has arrived, that sonochemistry is expanding 1991, 113, 2785.
and that chemists from all disciplines will find within the 23 A. Durant, J. L. Delplancke, R. Winand and J. Reisse, Tetrahedron Lett.,
subject plenty that will be of interest to them. 1995, 36, 4257.
24 J. Bujons, R. Guajardo and K. S. Kyler, J. Am. Chem. Soc., 1988, 110,
7 References 604.
25 K. Tadasa, Y. Yamamoto, I. Shimoda and H. Kayahara, J. Fac. Agric.
1 T. J. Mason and J. P. Lorimer, Sonochemistry, Theory, Applications and Shinshu Univ., 1990, 26, 21.
Uses of Ultrasound in Chemistry, Ellis Horwood Publishers, Chichester, 26 R. Bar, Biotechnol. Bioeng., 1988, 332, 655.
1988. 27 R. S. Davidson, Ultrasonics, 1987, 25, 35.
2 Ultrasound, its physical, biological and chemical effects, ed. K. S. 28 S. L. Regen and A. Singh, J. Org. Chem., 1982, 47, 1587.
Suslick, VCH, Mannheim, 1988. 29 T. Hanafusa, J. Ichihara and T. Ashida, Chem. Lett., 1987, 687.
3 C. Einhorn, J. Einhorn and J.-L. Luche, Synthesis—Stuttgart, 1989, 30 C. M. R. Low, Ultrasonics Sonochemistry, 1995, 2, 153.
787. 31 C. Yang, D. T. C. Yang and R. G. Harvey, Syn Lett., 1992, 799.
4 Sonochemistry: The uses of ultrasound in chemistry, ed. T. J. Mason, 32 G. E. Keck, A. Palani and S. F. McHardy, J. Org. Chem., 1994, 59,
Royal Society of Chemistry, Cambridge, 1990. 3113.
5 Current trends in sonochemistry, ed. G. J. Price, Royal Society of 33 C. Garot, T. Javed, T. J. Mason, J. L. Turner and J. W. Cooper, Bulletin
Chemistry, Cambridge, 1993. des Societes Chimiques Belges, 1996, 105, 755.
6 Special edition of the journal Ultrasonics covering the RSC Sonochem- 34 T. J. Mason and J.-L. Luche, Ultrasound as a new tool for synthetic
istry Symposium, Warwick 1986, Ultrasonics, 1987, 25, January chemists, Chemistry under Extreme or Non-classical Conditions, ed. R.
issue. van Eldick and C. D. Hubbard, John Wiley, New York, 1997, p. 317.
7 T. J. Mason, Practical Sonochemistry, A users guide to applications in 35 J. Reisse, presented at NATO Advanced Study Institute on Sonochem-
chemistry and chemical engineering, Ellis Horwood Publishers, Chich- istry and Sonoluminescence, Leavenworth, Washington, USA, August
ester, 1991. 1997.
8 J.-L. Luche, Sonochemistry, from experiment to theoretical con-
siderations, Advances in Sonochemistry, ed. T. J. Mason, JAI Press, Received, 9th May 1997
London, 1993, vol. 3, p. 85. Accepted, 30th June 1997
Chemical Society Reviews, 1997, volume 26 451